The following explains the structure of the data and programs to estimate results reported in "CONSUMER INERTIA, CHOICE 
DEPENDENCE AND LEARNING FROM EXPERIENCE IN A REPEATED DECISION PROBLEM" and the corresponding appendix.

Unzip RSTATDAT.ZIP and then run in Gauss the following two files in this order:

RESTAT1.DEF   - Reads original data.
              - Simulates a choice based dataset.
              - Computes everything (and more) reported in the paper except the Arellano-Carrasco estimator.
              - Saves output to RESTAT1.out
              - Saves Gauss files exogenas, knames, rep, and dnames.

RESTAT2.DEF   - Computes all histories and choice probabilities at each node.
              - Estimates each of the GMM estimates of Tables 5 and 6 as well as those for different subsamples defined
                by demographics (not reported in the paper).
              - Computes marginal effects of Table 7.